U.S. Clinical Specialist – Visual Question Answering (VQA) Project (V00273)

We are seeking U.S.-based, board-certified clinicians and imaging specialists to generate high-quality Visual Question Answering (VQA) content grounded in real clinical practice. The objective is to create rigorous, domain-informed Q&A pairs based on real clinical visuals—charts, dashboards, flowsheets, tables, diagnostic images (CT/MRI/X-ray), pathology slides, vitals trends, ECGs, and other specialty-specific artifacts used in everyday care.

You’ll provide fully de-identified or publicly available visuals from your day-to-day workflow and transform them into precise, reasoning-heavy Q&A pairs that reflect authentic clinical interpretation. This dataset will train advanced LLMs to handle specialty-level reasoning, including image-based logic, clinical inference, cross-panel comparison, and guideline-aware decision-making.

• Eligibility: U.S.-based, board-certified physicians, nurse practitioners, registered nurses, and imaging technologists (radiology or cardiology).
• Availability: Steady weekly participation required; approximately 10–15 hours per week (flexible) and 2 hours per day.
• Mode: Asynchronous; work completed through V7 Go, a secure, web-based annotation and review platform.
• This project focuses on transforming real clinical artifacts—such as imaging studies, pathology materials, vitals trends, ECGs, flowsheets, dashboards, and other specialty-specific visuals—into high-difficulty Q&A pairs that demonstrate authentic clinical reasoning. Each submission should reflect the interpretation, judgment, and domain expertise clinicians apply in everyday practice.

• Provide fully de-identified or publicly available clinical visuals sourced from your day-to-day practice.
• Produce high-difficulty, domain-grounded questions requiring interpretation, reasoning, and specialty-level judgment.
• Write clear, complete, and logically justified answers grounded explicitly in the visual.
• Ensure each Q&A pair reflects authentic clinical reasoning, not simple data extraction.
• Adjust question complexity based on LLM performance and QA feedback.
• Maintain consistency, accuracy, and adherence to project guidelines across all submissions.

Follow privacy rules, documentation standards, and platform instructions within V7 Go.

• U.S.-based and actively licensed in one of the following roles:
• • Board-certified physicians: Radiology, Thoracic Radiology, Neuroradiology, Imaging Cardiology, Pathology, Primary Care, Internal Medicine, Pediatrics
  • Nursing: Nurse Practitioners, Registered Nurses
  • Imaging professionals: Radiology Technologists or Cardiology Technologists.
• Capacity to contribute 10–15 hours per week, flexible depending on pace and specialty, ensuring consistent weekly participation. Minimum commitment per day: 2 hours. 
• Project duration: approx. 1 month. 

Comfortable with providing fully de-identified or publicly available clinical visuals from routine workflow.